Parry...

[Menae]

So, Casper is super duper correct. Like, super duper.

For one, let's just say the op's estimate of like 1-2% damage reduction from full parry was right. Yes, this sounds trivial, but you know what else only reduces about 2% damage over time? Hallowed Ground.

Rampart: 20 seconds every 90, 20% reduction = 4.44% DR over time.
Sentinel: 10 seconds every 180, 40% reduction = 2.2% DR over time.
Hallowed Ground: 10 seconds every 420, 100% reduction = 2.38% DR over time.
Convalescence: 20 seconds every 120, 30% healing received = 5% HR over time.
Total increases from cool downs:
9.04% DR over time
11.58% eHP over time
16.58% HR over time

This is a fallacious argument. You are conflating damage taken over the course of a fight with the damage taken during a very specific portion of the fight. Hallowed Ground is valuable not because of it's overall damage reduction, but its ability to completely ignore specific attacks or dangerous portions of the fight. It allows healers to reposition or heal the raid or top you off. Parry does none of these things; it merely provides a chance for you to take reduced damage from an attack, which may or may not fire off, meaning it is inconsistent. Parry's real value is hopefully allowing your healers to rely more on their smaller, more efficient heals by removing their tendency to freak out when you take a big damage spike via reducing the occurance of those spikes. It CAN save your life, no doubt - but if the only reason you lived through an attack was because you parried it, it's no different from the only reason you living through an attack being the boss rolled low on damage - it's luck, and you should not count on it.

So, if full parry gave you 2%, it would be equal to having an extra sentinel to pop over time. More, this would amount to a 20% increase in total dr over time (2% added to the ~10% from abilities.) so yeah.... It's pretty cool even at this low value.

Not to mention my own testing on turn 5 pre 2.2 has agreed with Casper's value. The returns are very high.

Where parry shines is on big hitting mobs in very strenuous situations. Having so much parry made me able to solo tank twin, which was the only way my group could win. Continuing to have it is allowing us to solo tanking turn 6. Those percentages add up. In places like 5 and 6, in many ways the auto attacks are more dangerous. You don't get killed by the 8 stack cleave, it's the AA after it that drops you. High parry and block gives you a strong statistical chance to mitigate and live in danger situations.

Ps: @50 parry = 1% dr, I'm at about 620 atm. Aka 12.4% Aka more than my entire cool down suite.

This is exactly the WRONG situation. Parry is most valuable in situations where you have lots of smaller hits incoming so that bad luck is less likely. Think about it this way: you have 50% parry (whatever number that requires) and 50% block, both at 25% DR. The boss has a two hit attack coming at you that does a combined 110% of your remaining health (after any CDs/etc you put up); if both connect you die, but even one parry/block saves you. You die 25% of the time. If that attack is instead 10 hits, say, from 10 smaller mobs instead of a large boss, then you will die less than 2% of the time. Parry is also good in any situation where a single hit will barely kill you, as it provides a chance to live where any other stat might not. That said, the best solution in those situations is usually get a heal.

That's not to say parry can't save your bacon - we've all had times where the tank drops to 50 hp and said, whew, thank god I blocked/parried THAT. I'm not necessarily saying parry is a bad stat; I'm saying it's not good for the reasons you are saying it is. The reason you can solo tank T6 isn't because parry is saving you - if it were, you'd be reliant entirely on luck to survive, and that's terrible tanking strategy. Rather, the reason you are solo tanking T6 is because you and your healers are skilled enough and geared enough in NON-RANDOM stats so that they can consistently keep you alive through your cooldowns and the increasing stacks of the boss.

[subteraneanbird]

As an up and coming tank, this thread is interesting. I do agree that testing so far has been inconclusive but at the same time, since no one really seems to have any solid idea on the exact returns on Parry, it's something I'm gonna keep in mind.

Also comparing what is essentially an RNG factor to something that is highly consistent like a CD is awful. You can never get a parry at all for a fight and then its effectively 0%. You can parry a string of big hits and then its much higher. What separates a CD and parry is you know exactly when a CD is gonna proc, because you're the one setting it. Even ones that can have an effective reduction rate of 0% (looking at you, Featherfoot and Awareness) are better than parry because you can time them more effectively than just waiting for a lucky parry proc. Parry by its nature is inconsistent, and that's okay. But pretending that it isn't, and is comparable to highly consistent skills that form part of the backbone of tanking is dishonest.

[PiedPiper]

You're straw manning my point, and yes you're right that it's not a one to one comparison. Of course cool downs are different and valuable because of their precise nature.

I was just showing that 2% dr over time isn't as small as it seems.


More, because Casper's numbers are right (or at least right-er), parry can be pushed high enough that you can begin to "count" on it over time.

By stacking a stat like parry, favoring buckler over kite shields and so on, the cumulative effect if these stats work together to greatly flatten the damage you receive over time.

[PiedPiper]

More to the point, this thread is discussing parry vs dps stats and (OP) saying the odds of parry saving you (thus preventing a wipe) are lower than the odds of the tank's dps increase making the entire group pass some dps check (thus preventing a wipe).

As Casper pointed out. A 20% dps increase on a paladin is still only about 20 dps, in a group of 4 dps, all of which should now pull no less than 200, that is adding 20 dps to about 1000 dps group total (counting both tanks) give or take a few points. This is a VERY small increase.


Parry, however, at the 600+ numbers imparts about a 45-50% chance to parry an attack you fail to block (which you already stood a 40ish% chance to block with a buckler). This means every attack you take has to (nearly) flip and land on tails twice to take full damage. This is actually pretty reliable. Even over the seconds they are active, foresight and bulwark are both weaker than highly stated parry.

[PiedPiper]

You guys seem to thing cus it's random and has a small chance it's unreliable. On one hit, this is true (though you are more likely to block or parry than to fail to mitigate). The fights where any of this is relevant are all long fights. It will have good returns over time. ~45-50% isn't a small chance.


Which brings us back to high damage situations like solo tanking 5, 6, or 8. Sustained damage in these situations is so high, that at just about any moment, you could go down. (Even in full 2.0 bis twin would still occasionally hit me for 7000, cus and that's just bad luck.) if you get unlucky, no matter what you are wearing, you will die. But, if you wear 100 more parry, you will get that unlucky about 8% less.

How often is your 20dps gonna save the raid? More often than 8 in 100?

Is 20 dps worth loosing any amount of chance to cling to life and continue tanking?

100 secondary stats = 8% increase chance to parry, or 2% increase to group dps (at best).

[PiedPiper]

Also comparing what is essentially an RNG factor to something that is highly consistent like a CD is awful. You can never get a parry at all for a fight and then its effectively 0%. You can parry a string of big hits and then its much higher. What separates a CD and parry is you know exactly when a CD is gonna proc, because you're the one setting it. Even ones that can have an effective reduction rate of 0% (looking at you, Featherfoot and Awareness) are better than parry because you can time them more effectively than just waiting for a lucky parry proc. Parry by its nature is inconsistent, and that's okay. But pretending that it isn't, and is comparable to highly consistent skills that form part of the backbone of tanking is dishonest

And, again, with very high parry stats, your chance to parry becomes high enough that you can begin to depend on it and block to work together to save you with some frequency.

[PiedPiper]

No good tank activates Rampart whenever it's up, nor Sentinel, AND NEVER HALLOWED GROUND.

Counter Example: in turn 5, you do actually use all of them pretty much on cool down. The way death sentence timing works out is that if you use ramp, forsight + conv, sent, ramp is up for next, fs+conv, bulwark, ramp, fs+conv, sent etc.


So yeah. Sometimes the best use is almost on cool down. Oh, and with parry on top of your cool downs, sometimes death sentence hits you for exactly zero, and that's pretty cool.

[ZDamned]

No, You made a terrible point. Flat Out, there is no comparison.

You failed to prove anything, because 2% dr IS very small over time. Were still looking at sub 5000 numbers here. Thats less than a full HP Bar for a Paladin.
AND Unless your party has no Bards in it, chances are, your healers are perfectly fine on MP.

If that was true, then why are we losing parry going into "BIS" ilvl110? Does SE want us to take more damage?

There is not enough points to allot in any 2ndary stat to make anything become reliable enough to count on except for the raw DPS increase of Determination.

Even if you look at it over time, the encounters still vary too much, and the damage reduced over the course of a fight is honestly negligible.

If you are in a situation where your Health is consistently hovering at a range where you will die unless you parry the next hit, then your healers are not good enough for the encounter you are on.

Ever since I swapped over to Crit/Determination gear, my healers have still said they noticed no difference at all in healing me. If the healers out there don't care, then why should we reduce our damage output?
We also become more successful in Downing T6 AFTER I changed, so obviously the BIS 90 was not exactly making or breaking anything at all even while learning a new encounter.

Counter Example: in turn 5, you do actually use all of them pretty much on cool down. The way death sentence timing works out is that if you use ramp, forsight + conv, sent, ramp is up for next, fs+conv, bulwark, ramp, fs+conv, sent etc.


So yeah. Sometimes the best use is almost on cool down. Oh, and with parry on top of your cool downs, sometimes death sentence hits you for exactly zero, and that's pretty cool.

No, Just because that is how you tank T5, doesn't mean that is what everyone does. I solo tank T5, and use my CD's for the initial pull, then I save my CD's for Snakes and for Healer locks during Liquid Hell. Hallowed Ground is ONLY used when I get targeted by Liquid Hell at the end. Blowing CD's just for Death Sentence means your Healers are bad, and can't learn how to time Stone Skin/Adlo.

Parry, however, at the 600+ numbers imparts about a 45-50% chance to parry an attack you fail to block (which you already stood a 40ish% chance to block with a buckler). This means every attack you take has to (nearly) flip and land on tails twice to take full damage. This is actually pretty reliable. Even over the seconds they are active, foresight and bulwark are both weaker than highly stated parry.

You are so wrong... If I could parry 45-50% of the time, I would never have started this thread.
PS: I have more parry than you do, and I NEVER see your magic numbers.

[subteraneanbird]

There is no amount of parry that will let me say "I can parry this Death Sentence (or whatever other huge hit move) and not all these mostly meaningless auto-attacks." That's the issue. It will always be inconsistent because it is an RNG stat. It is a stat that won't even come into play 100% of the time because of how it works with positioning. It is not possible to will yourself into parrying any one specific hit, while you can definitely pop a CD for a specific hit. If parry were able to be pushed high enough to be that effective, then it would be exceedingly obvious wouldn't it?

I think S-E learned from XI that mitigation based tanking is hell. No one likes being wholly dependent on a lucky die roll to save their ass.

[PiedPiper]

You are so wrong... If I could parry 45-50% of the time, I would never have started this thread.
PS: I have more parry than you do, and I NEVER see your magic numbers.

You can't have much more parry than me lol. I'm over 620 with food, it's simple not possible to be much more.


That said, of course you aren't parrying 45%, not all attacks can be mitigated and you block then parry. Of the parriable attacks that you don't block? Yeah you are parrying 40%+. Do you play with floating text on?



Also, I was taking about cding in final phase not before that.